WRF wind forecast over coastal complex terrain: Baja California Peninsula (Mexico) case study

Weather Research and Forecasting (WRF) near surface wind forecast sensitivity to planetary boundary layer (PBL) scheme over complex terrain of Baja California Peninsula, México, is examined. Yonsei University (YSU), Mellor-Yamada-Janjic (MYJ), and Asymmetric Convective Model version 2 (ACM2) PBL schemes are evaluated using the Taylor diagram, mean absolute error skill score (MAESS), and mean absolute error standardized anomaly metrics. Additionally, forecasted wind ramp distribution is analyzed. YSU scheme improves forecast accuracy in winter for most of the weather stations. Meanwhile, during summer, the performance of PBL schemes varies depending on physiographical environment of the weather station site. WRF forecast tends to generate a greater number of up/down wind ramp events than observed in the range of 2 m/s. The diurnal behavior of wind speed is well reproduced by all PBL schemes; however, the wind speed variability is smoother than the observed. The ability of the ACM2 scheme to perform well in winter and summer may be related to the critical factor that determines the contribution ratio of non-local mixing to total turbulent mixing. The WRF is capable of accurately forecasting the synoptic-scale energy power spectrum in winter; however, in the mesoscale range, the simulated spectrum underestimates the energy for both seasons.